prehisle/relay-pulse

Health v2 · maintenance only

Adoption not part of the score

1098 stars · 93 forks observed · 2026-09-03

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

RelayPulse is a self-hosted monitoring service for LLM API relay/proxy services that probes endpoints with real token-consuming API calls and validates responses, avoiding false 'HTTP 200 but empty output' uptime. It provides a web dashboard with availability heatmaps (24h/7d/30d), hot-reloadable config, SQLite/PostgreSQL storage, and Telegram/QQ bot notifications.

Use cases

• monitor availability of my LLM API relay service
• detect when an LLM API returns 200 but empty responses
• compare latency and error rates across multiple LLM providers
• track uptime of self-hosted OpenAI-compatible endpoints
• get notified when an LLM API goes down
• visualize API service quality with an uptime heatmap

When to choose

• you need real-call validation of LLM APIs rather than simple HTTP ping checks
• you run or buy LLM relay/proxy services and need continuous availability tracking
• you want a self-hosted, cloud-native monitor with Docker deployment and hot config reload

When to avoid

• you only need basic HTTP endpoint uptime checks without token cost
• you cannot expose API keys or spend tokens on periodic probes
• you need monitoring of non-LLM services only

Facets

service · maturity active

monitoring alerting http-client data-visualization self-hosted monitoring large-language-models apis self-hosted self-hosted go llm-api-monitoring uptime-monitoring synthetic-checks status-dashboard api-relay devops docker linux web-server

Member repositories

For agents

markdown · JSON · MCP: product_card(name="prehisle/relay-pulse")